JP2003281561A - Map data processing method, and computer graphics processing method and device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a front determining function-applied computer graphics processing technique capable of determining the front of a building in a map. <P>SOLUTION: The computer graphics processor 10 has map data processing function for performing a front determination. Namely, a front determination part 18 determines, from map data acquired by a map data acquisition part 12, the front of a building including in map data on the basis of the distance form a ambient building. The face having the largest distance to the ambient building is determined as the front of the building. The determination processing is simplified by using a bounding box for surrounding the building. The computer graphics processor 10 generates a three-dimensional computer graphic image including the building the front of which is expressed according to the front determination result. Concretely, a texture generation part 20 pastes a texture showing an entrance to the front of the building. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a map data processing method and a computer graphic processing method and apparatus using the map data processing method.

[0002]

2. Description of the Related Art In recent years, the development of computer graphics (hereinafter, CG) technology has been remarkable, and it has become familiar because of many opportunities for us to see it. CG has made a great contribution to new fields such as movies and games. There is no doubt about the persuasive power of the three-dimensional display. The number of active fields for CG will continue to increase. If you look at the current CG world, you will find that C
There are not so many G's. But,
The situation in which one person creates a CG for himself, from the situation in which experts create a CG, just as it became a pleasure to take a picture from a passive position of having a photograph taken Will change to. From such a point of view, there is a problem that enormous cost and time and the technical skill of the creator are required for the current CG creation. As a technique to solve it, image-based modeling (eg, to create individual objects from photographs)
3D CG made from photos and photos-Image-based modeling & rendering-2001)
It is a very effective technique for creating G. You can create the desired 3D object even if you do not have much knowledge about 3D modelers. However, in order to reproduce the streetscape with many buildings, it takes a lot of time and effort to create objects individually. When creating a cityscape, the problem is how much more objects can be created easily than the degree of perfection of each object.

Currently, as a technique for generating a three-dimensional cityscape from a digital map, one that dynamically generates textures using a digital map and an aerial photograph (use for a three-dimensional GIS and a fire support system, http: // www.isad.or.jp/magzin/61_3z
igen.htm), which generates 3D using only digital maps (Kenichi Sugihara, 3D automatic streetscape generation system by integrating GIS and CG, http://www.gis.pasco.co.jp/c
ommunity / doc_lib / uc2000 / user_pdf / p2000_4..pdf) is a typical method. When using aerial photographs, the degree of completion is certainly improved, but the time taken to create them cannot be expected to decrease significantly, and it is difficult to say that it is easy. Also, an aerial photograph that matches the location of the digital map to be created is required, and the method of obtaining the data is by no means simple. However, since the GSI started the aerial photo browsing service on the Internet in April 2002, there is a possibility that this method will be effective. However, this method is not suitable for obtaining ease and speed. In the case of another method, that is, only creating a digital map, how to prepare the texture becomes a big problem.

The present inventor conducts research aiming at convenience and speed while considering the above problems. Then, the present inventor conducts research on graphic processing in which a three-dimensional cityscape is generated without using an aerial photograph and a texture is dynamically generated.

In the research, while utilizing map data of a cityscape including buildings, the present inventor has noticed that such map data does not have information on the front of the building. So far, no technique has been proposed for determining the front of a building in map data. If the front of the building could be determined, such information would be useful for graphics processing. For example, it is considered that a more realistic image can be created by considering the front of the building.

[0006] With respect to the above points, the conventional high-level CG creation technique generally targets modeling of a fine shape including the entrance of a building. That is, the front of the building is determined at the modeling stage. Therefore, there was no need to determine where the front of the building was. On the other hand, in a simple CG creation technique, processing based on information indicating where the front of a building is is not generally performed. Therefore,
Again, there was no need to determine where the front of the building was. Against this background, the inventor focused on the usefulness of determining the front of a building while researching a technique for easily creating a three-dimensional graphics image from map data.

Further, the present inventor, as described above, paid attention to the usefulness of the information on the front of the building in the research of graphics. However, the information on the front of the building is considered to be useful for other uses of map data such as directions. In this respect, the scope of the invention need not be limited to graphics technology.

The present invention has been made under the above background, and an object thereof is to provide a suitable technique for determining the front of a building in map data.

[0009]

The map data processing method of the present invention determines the front of a building included in the map data based on the distance to the surrounding buildings. Preferably, the map data processing method of the present invention determines that the surface having the largest distance to the surrounding building is the front surface of the building.

The present invention works particularly well when processing map data of a cityscape. The map data of the cityscape is
Typically, it is a map of the city, where there are relatively many buildings. The cityscape map may be an actual cityscape map or a virtual cityscape map. In a cityscape, generally, the distance between buildings across a road is relatively large, and the distance between buildings across a road is relatively small. And the front of the building often faces the road. Further, when a building faces multiple roads, it is often the case that the building faces a wider road. Therefore, the front of a building is generally a surface with a large distance to the surrounding buildings. According to the present invention, the front of the building is appropriately determined by determining the front of the building based on the relationship between the distance between the buildings and the front of the building.

The map data processing method of the present invention is typically applied to a computer graphics processing technique described later. However, the present invention is not limited to this. The front determination can be applied to other uses of map data. For example, the technique of the present invention may be applied to route guidance using map data. Further, for example, when drawing a road on a two-dimensional map of only buildings, the front determination result of the present invention can be preferably used.

Preferably, the map data processing method of the present invention determines the front of the building by using the distance between the bounding boxes surrounding the building as the distance between the buildings. According to the present invention, even when a building has a complicated shape, the front of the building can be obtained by a relatively simple process. The bounding box is typically square. In this case, even when the building has a shape other than a simple square,
Since it is only necessary to determine the front of the building from the distance between the quadrangles, the front determination processing becomes simple.

Preferably, in the map data processing method of the present invention, the distance from one bounding box to another bounding box is extended until the extension line of the side of the other bounding box reaches the one bounding box. The length of the line. The length of this extension line can be obtained by a relatively simple process. Therefore,
According to the present invention, the front of a building can be obtained by a relatively simple process.

Preferably, in the map data processing method of the present invention, a surround box corresponding to each side of the bounding box is formed, an intersection of the surround box and the bounding box is obtained, and a vertex of the bounding box included in the surround box is determined. A certain inclusion point is obtained, and the distance from the intersection point on the side of the surround box to the inclusion point is used for the front determination as the distance between the bounding boxes. The above-mentioned intersection points and inclusion points are relatively easily obtained. Therefore, according to the present invention, the front of a building can be obtained by a relatively simple process.

It should be noted that the surround box is typically polygonal, and preferably rectangular. In the embodiments described below, the surround box is rectangular. When performing processing with rectangles, in special cases,
Two adjacent sides of a rectangle have the same length, resulting in a square, but the square thus created may be considered to be included in the rectangle in the present invention.

Preferably, when there is no other building around the building, it is determined that the largest face of the building is the front surface. For example, in the embodiment using the rectangle described above, the rectangle may not intersect with other bounding boxes. In such cases, the largest face of the building is often the front. Therefore, according to the present invention, the front of the building can be appropriately determined.

According to another aspect of the present invention, there is provided a map data processing method for carrying out a front face judging process for judging that the largest face of a building included in the map data is the front face of the building. The front of a building is often the largest of the multiple surfaces that a building has. Therefore, according to the present invention, the front of the building can be appropriately determined.

Another aspect of the present invention is a computer graphics processing method, which is based on the step of obtaining map data and the front of the building included in the map data based on the distance to the surrounding buildings. And a computer graphic image is generated from the map data using the determination result of the front surface. This aspect is
A computer graphics image is generated using the map data processing method described above. An appropriate computer graphics image is generated by using the determination result of the front of the map. For example, the texture of the doorway of the building is attached to the front of the building, and an image closer to reality is obtained.

Another aspect of the present invention is a computer graphics processing apparatus, which comprises a map data acquisition unit for acquiring map data and a distance between a front of a building included in the map data and a surrounding building. And a front face determination unit that determines based on the above, and generates a computer graphic image from the map data using the front face determination result. The present invention also produces suitable computer graphics images.

Another aspect of the present invention is a computer graphic processing apparatus, which is a map data acquisition unit for acquiring map data including a two-dimensional shape and arrangement of a building, and information about the height of the building included in the map data. Including a height acquisition unit for acquiring the front surface of the building included in the map data, and a front surface determination unit for determining the front surface of the building based on the distance to the surrounding building obtained from the map data, and acquired by the height acquisition unit. A three-dimensional computer graphic image including a building having a three-dimensional shape obtained by pulling up the two-dimensional shape to the designated height, and the front of which is represented according to the determination result of the front determination unit is generated. According to the present invention, a three-dimensional computer graphics image can be easily generated from two-dimensional map data. Furthermore, it is possible to generate an image that is closer to reality and that represents the representation of the building.

[0021] Preferably, the computer graphic processing apparatus of the present invention performs a process of attaching a texture representing the front of the building to the surface determined to be the front of the building. The texture representing the front of the building is, for example, the texture of the doorway. According to the present invention, an image closer to reality can be generated.

The front face determination result can be effectively used in addition to the above-described texture pasting process. For example, using the tendency that the front of a building often faces the road,
Roads between buildings can be automatically generated.

In another aspect of the present invention, the map data processing method carries out a front face judging process for judging the front face of the building included in the map data based on the information on the space up to the surrounding buildings. Preferably, the method of the present invention determines that the surface with the largest space to the surrounding building is the front of the building. In this aspect, the space to the surrounding building includes the distance described above.

However, the space may be information other than the distance. The space may be an area. As an example, a line perpendicular to the wall surface is drawn from both ends of the wall surface of the building. The area of the polygon formed by the wall, the perpendicular lines and other buildings is calculated. The surface that maximizes this area is determined to be the front of the building. A bounding box may also be used for this space determination.

As described above, the space may be other than the distance, and may be expressed by the area, for example. However, using the distance as the space is advantageous because the processing is simple. Especially by using the bounding box,
Furthermore, by using an extension line, an intersection, an inclusion point, etc., it is advantageous that appropriate determination can be performed by a simple process.

The availability of space information, that is, it need not be limited to distance, is another aspect, for example:
The same applies to the aspect of the computer graphics processing device.

It is to be noted that any combination of the above constituent elements, and the expression of the present invention converted between a method, an apparatus, a server, a system, a computer program, a recording medium, etc. are also effective as an aspect of the present invention. .

[0028]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention will be further described below. In the following, first, the study “construction of a three-dimensional cityscape CG image using a digital map” by the present inventor will be described. Then, an embodiment of a computer graphic processing device to which the same research is applied will be described. The computer graphics processing device is advantageously incorporated with the frontal judgment function in the following research.

1. First, in this research, we describe a method of creating each 3D object that constructs a cityscape from a digital map and a simple texture. The purpose is to create a two-dimensional object located at the midpoint between the digital city and the virtual city. With this system, anyone can easily create a virtual space in which multiple building objects are lined up in a short time.

Although there are many systems that can display two-dimensional images in GIS even now, it is not possible to write out three-dimensional data, and a system that requires separate aerial photographs, etc. There are many forms that are not suitable for. In addition, it takes a great deal of time and labor to create an entire cityscape, because an engineer needs to have a deep knowledge and skill for a modeler to create it from scratch using a modeler. This method is also not suitable for the purpose of this study.

In this research, three-dimensional data is instantly created from a two-dimensional digital map, which face is in front of the building object is determined, and textures are automatically generated to "create a virtual space easily and in a short time." Achieve the goal of "building".

2. System Configuration and Process Flow Figure 1 shows the system configuration and process flow in this research. In this study, Geo is used to handle digital map data.
32-bit WINDOWS (registered trademark, the same applies hereinafter) in which various functions required for GIS called Base are condensed
The tool library of DLL is used. The system first reads the data from the two-dimensional digital map. This digital map must be in a format that GeoBase can read. After that, the height of the read graphic group is changed and a new object is added. Even at this stage, provisional three-dimensional display is possible by utilizing the function of GeoBase. After that, the two-dimensional data having the height information is written, and the process is transferred to another program.

In the cityscape creation program, height information and 2
Three-dimensional polygon data is generated from the dimensional data, and the distance between the objects is calculated to determine the front and side surfaces. Then, the texture is automatically generated and pasted. After that, each object is manually corrected and the process is completed. In addition, in this study, OpenG is displayed.
L is used.

3.2 Processing on Two-Dimensional Numerical Map 3.1 GeoBase Application and Basic Processing An image of GeoBase used in this application is shown in FIG. This application can read digital maps, add / delete building objects, and change the height. In addition, you can preview what the appearance will look like when it is processed into three dimensions in the state of a primitive object. After that, specify the range and write out two-dimensional vector data.

GeoBase is not an application but a group of libraries that assists in application creation. FIG. 3 shows the functions provided by GeoBase. Of the support functions shown in FIG. 3, the following five basic processes were created by using the four controls of window control, color palette control, vector drawing display, and figure operation command. That is, (1) reading of a numerical map, (2) zoom-in / zoom-out, (3) pick processing, (4) addition of graphics / color change, and (5) three-dimensional display. By using these functions, it is possible to change the two-dimensional map to the layout that one imagines.

Referring to FIG. 4, the right window is a two-dimensional map and the left window is a three-dimensional display. When an object addition process is added to the two-dimensional map, it is reflected in the left window in real time. Then this G
The height change using the mouse, which is the most characteristic of the eoBase application, will be described.

3.2 Height Change by Mouse Since the GeoBase itself does not have a function for changing the height with a mouse, the pick processing and the graphic data changing function are combined. Referring to FIG. 5, it is possible to change the height in real time by selecting an object whose height is to be changed and dragging the mouse in that state. Although there is no progress in FIG. 5, the height changes smoothly on the actual execution screen.

The two-dimensional vector data created by the processing so far is written in a file with "height information, color, vertex data" as a set. The three-dimensional cityscape creation program creates triangular polygon data based on the data.

4. Generation of three-dimensional space 4.1 Generation of three-dimensional data from two-dimensional data Section 4 describes a method of generating three-dimensional data from the two-dimensional data discharged by the GeoBase application. First, FIG. 6 shows a part of the written data.

In FIG. 6, the first three values are RGB
(Color data), and the fourth value “2” represents the height.
The last "4" indicates the number of vertices listed from the next row. So this data is a quadrangle. Since the digital map is represented by a huge map, the coordinates are very large. This value is read, then normalized and converted to a manageable number. In the next section, a method of dividing this two-dimensional coordinate into three-dimensional polygons will be described.

4.1.1 Generation of Side Polygons Since the shape of the object of the building is limited to flat sides this time, the generation of side polygons is less difficult than that of the top surface and is very easy. Is. You can create the image as if pulling the surface up.

Referring to FIG. 7, the coordinates of two vertices on one side are A (x1, y1), B (x2, y2), and the height is H.
Then, the triangular polygon on the side surface becomes as shown in the figure. Coordinates other than A and B are C: (x2, y2 + H), D:
(X1, y2 + H). That is, the triangular polygon on the side surface can be created by connecting the vertices of "A, B, C" and "A, C, D". By applying this method to the other three sides,
You can get the coordinates of a dimensional solid.

4.1.2 Generation of Top Surface Polygon Since the shape of the top surface is often a simple polygon, the triangulation becomes more complicated than the side surface. The polygon division algorithm is "Voronoi polygon and Delaunay triangulation",
Although there is "Deroni triangulation" etc., this time because it uses OpenGL for the drawing program, I used the tiling routine of the GLU library, which is one of the OpenGL libraries (Mark J. Kigar
d, OpenGL Programming for the X Window System (Japanese version), p58, 1997).

For performance reasons, OpenGL supports only the drawing of convex polygons, and concave polygons must be divided into small convex polygons. A GLU library routine for tiling is shown in FIG.

The actual usage of the routine of FIG. 8 will be described in order; static GLUtriangulatorObj * tobj = NULL; Declare a pointer for storing the tiling object; tobj = gluNewTess (); Create a new tiling object and create it earlier. To the specified pointer; then specify the callback; gluTessCallback (tobj, GLU_BEGIN, (void (__stdcall
*) (void)) glBegin); gluTessCallback (tobj, GLU_VERTEX, (void (__stdcall
*) (void)) GetVertex); gluTessCallback (tobj, GLU_EDGE_FLAG, (void (__stdcal
l *) (void)) GetFlag); gluTessCallback (tobj, GLU_END, (void (__stdcall *)
(void)) glEnd); Of these four specifications, GLU_VERTEX and GLU_EDGE are important. GLU_VERTEX specifies a callback routine that is called each time the vertices of a concave polygon registered by the gluTessVertex () routine are divided into triangles. Here it is called GetVertex, but this function is _stdcall vo
It is defined as id GetVertex (void * data) and receives the values of three GLdouble type vertices by a void type pointer. By holding it as a member variable, triangulation can be performed. However, with this callback GLU_EDGE
You cannot receive the correct value unless you also specify. The callback function here is _stdcall void
Although described as GetFlag (GLboolean b), this flag indicates whether the value of the vertex passed to GetVertex called next to this GetFlag is in the vertex of the concave polygon registered at the beginning. is there. We don't need this value in our process, but we must specify this callback to get the values of the correctly split vertices.

GluBeginPolygon (tobj); for (int i = 0; i <VerNum ++) {gluTessVertex (tobj, vertex [i], vertex [i]);} gluEndPolygon (tobj); Then tiling with gluBeginPolugon (tobj) And repeat the loop for the number of vertices in my polygon, gluTessV
Pass the vertex value in the form of float [3] or double [3] in ertex.
The second and third arguments may be the same. And finally gluE
By executing ndPolygon, the callback function specified earlier is called continuously, and the tiled vertices can be obtained.

4.2 Frontal Judgment In this research, various automatic processes are realized by automatically judging the frontal surface. The determination item uses the distance to the surrounding building (the surface with a long distance is the front) and the size of the surface. First, the former method will be described.

A rectangular parallelepiped bounding box is used to calculate the distance to the surrounding buildings. Create a rectangle with a constant vertical length for each of the four sides of the created bounding box, and calculate intersections with other sides of the rectangle and points included inside for all other building objects. . Calculate the distance to each object based on that value,
Used for frontal judgment. The bounding box will be described in the next section.

4.2.1 Bounding Box A bounding box is a collision determination area surrounding an object. The simplest creation method uses the maximum and minimum X and Y coordinates of a certain object, and although it can be created very easily, it has the drawback that the determination becomes ambiguous. For example, when making a bounding box as shown on the left side of FIG. 9, a large space is created at the lower right.

This time, this is for determining the space in front of the object, and this method of creating a bounding box is not suitable. Therefore, with the method described below,
Create a more bounding box to the object, as shown on the right side of FIG.

FIG. 10 shows a bounding box generation process. The feature of this method is that while the method of finding from the maximum and minimum values described above creates a side along the X and Y axes, the first side is the same as the largest side that symbolizes the object. It is to create with the inclination, and to create the next side based on that. Next, the created side is examined in the vertical and horizontal directions. Then, if it is vertical, check whether it is facing upward or downward, and if it is horizontal, check whether it is facing right or left. This is to determine which point the edge to be created next passes to form an edge that wraps the object.

A specific example will be described with reference to FIG. (FIG. 11, first row) First, the longest side of the object is searched for, and a side A parallel to it is created. The end point of this side A has not been decided yet. (FIG. 11, second row) Next, the direction of the created side A is examined. This time, the orientation is horizontal / right, so the next edge to be created passes through the maximum X coordinate and goes straight to the created edge. The side B in the second row in FIG. 11 is created using the conditions. Then, the intersection of the side B and the side A is obtained, and this is the end point of the side A. Next, the same determination is performed with reference to this side B. (Fig. 11, 3
Line B is vertical / upward. Therefore, the side to be created next passes through the maximum Y coordinate, so the side C in the third row in FIG. 11 is drawn. Similar to the above, the intersection point is obtained and set to the end point of the side B. (FIG. 11, 4th row) By applying the same method to the 4th side, a bounding box making use of the longest side could be created. Although there is still a space in the upper left part, assuming that it is a building, that part is on the premises and considered to be a part of the building object, which is rather convenient. By applying this algorithm, various bounding boxes can be created.

4.2.2 Space determination around building object As a concrete method of space calculation around an object,
Create a rectangle that is closely attached to the bounding box described above. The rectangle (hereafter called the surround box)
, And the intersection with the bounding box of another building object, and the distance from that intersection to your bounding box. Here, the above-mentioned intersection point and the vertex of the bounding box inside the rectangle (hereinafter referred to as the inclusion point) are obtained, and the distance to the obtained point is obtained. This distance is the distance from one side of the building object to another building object.

FIG. 12 is an example of the result of obtaining the intersection and the inclusion point, which is calculated for two building objects A and B. The four rectangles (A1 to A4, B1 to B4) provided around the rectangular object in the center are surround boxes. Point a is a surround box A3 (extends from bounding box A to the right)
And the bounding box B. Since the building object in this figure has almost the same shape as the bounding box, the intersection is on the side of the building object. In the surround box B3 extending leftward from the building (B) on the right side, the inner inclusion point b is pointed. The inclusion point b is the vertex of the bounding box A.

When there are a plurality of intersections / inclusion points on a surface,
Use the point with the shortest distance. For example, when two sets of intersections and inclusion points are obtained for a certain surface (one side of the bounding box), the distances between the intersections and inclusion points are compared, and the intersection and inclusion point of the combination having the shorter distance are adopted. . Thus, only one intersection corresponds to each face. Then, the distances of the adopted intersections and inclusion points are used as the space in front of the corresponding surface. This treatment is applied to all four sides of the bounding box.

From the above, the intersection points and the corresponding inclusion points on each surface of the bounding box are obtained. next,
The front determination condition using the intersection and the inclusion point will be described. In the front determination, (1) when there is no intersection: the front is determined by the size of the surface. (2) When there is one intersection: The surface with the intersection is the front. (3) When there are a plurality of intersections: Among the plurality of intersections, the surface having the longest distance to the intersection is the front. That is, the distance from the intersection on each surface to the corresponding inclusion point is obtained, and these distances are compared.
It is determined that the surface on which the intersection with the maximum distance is on the front. In detail, the face of the building corresponding to the calculated edge of the bounding box is the front of the building.

Next, a method of forming a surround box (rectangle extending around the bounding box) necessary for obtaining the intersection will be described with reference to FIG.

In order to fairly judge all building objects, the rectangle extending from the side of the building object is required to have a constant length and an inclination suitable for the object. In this program, the distance setting is 20. The following processing is performed in order to obtain a rectangle by obtaining a point with a constant distance from a side with a non-constant slope.

As shown in FIG. 13, a square having a predetermined length (20) is first prepared. Next, the two sides of the square are transformed into the lengths of the faces (sides of the bounding box) of the target building object. After that, the angle between the plane and the X axis is obtained using the inner product, and the rectangle deformed by that angle is rotated. Then move the created rectangle to the coordinates of the surface you want to closely contact. By the above processing, a rectangle with a certain distance can be created that is in close contact with the surface of the building object.

It is possible to determine a front face by calculating intersection points / inclusive points of the created surround box and all other building objects.

Here, it is easy to find the intersections by finding the intersections of straight lines. The inclusion point is determined as follows.

Referring to FIG. 14, in order to determine whether or not a point is included in a certain quadrangle, an outer product is used so that the point is in the same direction with respect to all sides of the quadrangle (this time, the surround box is left). Check whether it is located on the left side because it is defined around. If all are in the same direction, it is included, and if there is even one direction, it is outside.

The inclinations of the front surface of the bounding box determined by the above calculation and the surfaces forming the building object are compared, and the surface having the shortest angle and the shortest distance is set as the front surface, and the texture is changed.

4.3 Texture Generation Since there are building objects of various sizes and heights in the digital map and the height can be dynamically changed, it is not possible to prepare textures in advance. Therefore, it is necessary to dynamically generate many textures for various building objects. This time, I will describe a method that focuses on buildings.

The building wall design is formed in a monotonous pattern. It is also clear when looking at the real building. Therefore, by preparing the walls and windows separately and combining them so that the windows overlap with the walls, it is possible to create "wall type x window type" wall patterns. FIG. 15 shows a part of the prepared texture image.

By cutting out the white portion of the window and combining it with the wall, a pattern at one place can be created. By changing the number of times this is repeated according to the length and height of the surface, it is possible to create a texture suitable for the size of the building.
The number of times the texture was repeated was changed using OpenGL. FIG. 16 is an example of the execution.

5. Execution result 5.1 Experimental environment Dimension4300 from DELL was used as the environment for creating this system. CPU is Dual Pentium (registered trademark)
IV 1.7GHz, Memory 256MB, O
S uses Windows (registered trademark) 2000 and V
It was built with isual C ++ 6.0. OpenGL is used for the 3D graphics display library.
The operating environment can be executed by Windows 9x, Windows NT (registered trademark), and Windows 2000 that have introduced the OpenGL runtime library.

5.2 Results of Execution FIGS. 17 and 18 show the data created from the same numerical map. By the method of automatic texture generation, it was very easy to create a space where many such building objects exist.

Further, FIG. 19 is a diagram formally drawing the result of the front face determination. In FIG. 19, the pointed surface of the quadrangle is determined to be the front surface. The judgment of the front of one building object is different from other objects, but the reason is that there is another building object outside the drawing, and the intersection with it is far, that is, the spatial space is large. Therefore, the determination of the front is different from that of the building objects lined up horizontally. But,
Other results are almost as expected. From this, it can be said that the method of frontal judgment this time is practical. By adding an algorithm, it is possible to eliminate the fact that the front direction is different this time.

6. In this research, we examined a method to construct a three-dimensional space quickly and easily from a digital map. A large cost is required to manually construct an object that composes a cityscape. However, in recent years, opportunities to create virtual spaces such as 3D chat and games are increasing, and it is expected that individuals will create CG, and the need and demand for easily constructing 3D spaces are increasing. is there. In order to respond to this, in this research, we performed automatic generation of a large number of building objects by dynamic generation of textures, and judgment of the front that can be applied to the creation of entrances and roads. Also, this time, I used a digital map with only simple information. This is because in the future, we are considering a system that can create a cityscape without using a digital map and GeoBase. By doing so, I think that it is really easy to construct a three-dimensional space.

Other references for this research are: Masafumi Uehara, Jinbei Higashi, "Construction of 3D model of wide area city using digital map", IEICE Transactions D-II, Vol.J84D.
-II, No.8 pp1921-1924, 2001 ".

[Embodiment of Computer Graphics Processing Apparatus] An embodiment of the computer graphics processing apparatus to which the invention of the above research is applied will be described below.

FIG. 20 shows a computer graphics processing device of this embodiment. In terms of hardware, this configuration is the CPU, memory,
It can be realized by other LSIs, and in terms of software, it can be realized by a program loaded in the memory.
Here, the functional blocks realized by those collaborations are drawn. Therefore, it will be understood by those skilled in the art that these functional blocks can be realized in various forms by only hardware, only software, or a combination thereof.

The apparatus of FIG. 20 implements the map data processing method of the present invention, as described below. That is, this device has a function of determining the front of the building in the map data. Then, this device uses the front determination result to generate a three-dimensional computer graphics image (CG image) of the cityscape. The functions of this device are realized by the computer executing the program of the present invention.

Further, the apparatus of FIG. 20 is composed of a general-purpose personal computer. Therefore, input / output devices such as a keyboard, a pointing device, a display, and a printer are provided. Then, the generated image is displayed on the display. However, the present invention is not limited to this. For example, the device of the present invention may be provided with a WWW server and connected to the Internet. Then, this apparatus may receive a request from a user computer via the Internet, perform data processing in response to the request, and return the result to the user computer. At this time, a computer graphics image is output toward the user computer.

As shown in FIG. 20, the computer graphics processing device 10 includes a map data acquisition unit 12. The map data acquisition unit 12 acquires map data from a recording medium such as a hard disk device or a CD-ROM, for example. The map data acquisition unit 12 may acquire map data from the outside by communication.

The map data may be any data as long as it has information indicating the two-dimensional shape (shape viewed from above) and the layout of the building. The map data is typically the relatively simple digital map described above. However, the map data is not limited to this within the scope of the present invention. The map data may be a more detailed numerical map.

Further, the map data is typically an electronic map of an actual cityscape, but is not limited to this. The map data may be data such as a virtual cityscape. For example, the computer graphics processing device 10
Using, the user draws a group of buildings in the city on the screen. The building is drawn and the layout of the building is determined. The data thus created is acquired as map data by the CG creating function of the computer graphics processing device 10 itself.

The height information acquisition unit 14 acquires information on the height of the building included in the map data acquired by the map data acquisition unit 12. As described above, the user uses a pointing device such as a mouse to specify the height of each building. The designated height is acquired by the height information acquisition unit 14.

It should be noted that the configuration for acquiring the height information is not limited to the above, within the scope of the present invention. For example, the numerical value of the height may be input by the user. Also, a default height may be automatically given to the building and the height may be adjusted by the user as needed.

Further, in this embodiment, a CG of a part of the map data may be created. In this case, the user specifies a part of the map. Then, the height information of the buildings included in the specified range is acquired.

Next, the three-dimensional data generator 16 generates three-dimensional data based on the map data and the height information. The three-dimensional data generation unit 16 calculates the two-dimensional shape of the building
A three-dimensional shape is obtained by pulling up to that height. This processing is realized on the computer by generating the side surface polygon and the upper surface polygon by the processing described above.

The front judging section 18 judges the front of the building included in the map data based on the map data. The building front determination processing is as described above. That is,
The front of the building is determined based on the distance to the surrounding building obtained from the map data. The front surface determination unit 18 determines that the surface having the largest distance to the surrounding building is the front surface of the building. In this embodiment, a bounding box that surrounds the building is generated. Then, the distance between the bounding boxes is used as the distance between the buildings.

The texture generating section 20 carries out a process of attaching a texture to the wall surface of the building. As already mentioned,
Multiple textures are prepared in advance. One of them is selected for each building. This simplifies the user's work for texture generation.

The output processing section 22 performs a process for outputting a CG image to which a texture is attached. The output device is typically a display. At this time, the output processing unit 22 performs processing for displaying the CG image on the display. Further, when the output device is a printer, the output processing unit 22 performs processing for outputting print data to the printer. Further, when the CG image is sent to another computer, the output processing section 22 performs a process for sending the CG image data. Furthermore, the output processing unit 22
May perform a process for recording a CG image on a recording medium (including a hard disk) as one form of the output process.

FIG. 21 shows the structure of the front face judging section 18. The bounding box generation unit 30 performs the processing shown in FIGS. 10 and 11 to generate a bounding box that surrounds the building. The bounding box may be rectangular or rectangular parallelepiped. In the case of a rectangular parallelepiped, a plane-shaped rectangle is formed by the processing of FIGS. 10 and 11 and used in the processing described later. Rectangle generator 32
Performs the processing shown in FIGS. 12 and 13 to generate a rectangle (surround box) that is in close contact with the bounding box.

The bounding box and rectangle are
It is generated for each of the plurality of buildings included in the map data. Around each building, as shown in FIG.
Two rectangles are generated.

The intersection / inclusion point detecting unit 34 detects the intersection and the inclusion point using the bounding box and the rectangle. The meaning and detection process of intersections and inclusion points are
This has already been described with reference to FIG. The intersection is the intersection of the rectangle and the bounding box. The inclusion point is the vertex of the bounding box included in the rectangle. A combination of intersection points and inclusion points on one side of one rectangle (see FIG. 12) is used in the following processing.

One bounding box has four sides. An intersection and a corresponding inclusion point are obtained for each side. A rectangle closely attached to one side may have multiple inclusion points. At this time, there are a plurality of intersections on one side (there are a plurality of combinations of intersections and inclusion points). In such a situation, as described above, the intersection point that minimizes the distance between the intersection point and the inclusion point is selected.

The maximum surface determination unit 36 determines the maximum surface of the building. The maximum surface determination unit 36 determines which surface is the largest for each of the plurality of buildings included in the map data.

The front determining section 38 determines the front of the building. The front surface determination unit 38 determines which surface is the front surface for each of the plurality of buildings included in the map data. In the front determination, (1) when there is no intersection, the front is determined by the size of the surface. The front surface determination unit 38 determines that the largest surface is the front surface. (2) If there is only one intersection,
The front surface determination unit 38 determines that the surface having the intersection is the front surface. (3) When there are a plurality of intersections, the front determining unit 38 compares the distances between the intersections and the inclusion points. The intersection with the largest distance is required. The front surface determination unit 38 determines that the surface on which the obtained intersection is on is the front surface of the building.

In cases (2) and (3), the side of the bounding box is required as the front, not the actual surface of the building. Therefore, the front surface determining unit 38 obtains the face of the building corresponding to the obtained side of the bounding box.
The inclination and distance between the side of the bounding box and the surface of the building are used for this determination. A surface having a short distance and a small inclination is selected. In this algorithm, for example,
First, the side of interest (the side of the bounding box)
Building faces near the side opposite to are excluded (ie, building faces near the opposite side of the bounding box are excluded). Then, a building surface whose inclination with respect to the side of interest is equal to or less than a predetermined value is selected. Furthermore, the building surface closest to the side of interest is selected. This distance may be defined appropriately. This algorithm can be suitably applied even when a part of the building is bent. In this way, the front of the building is finally sought.

In the present embodiment, the sides of the two-dimensional shape of the building may be used as information indicating the direction and size of the surface of the building.

The correcting section 40 corrects the result of the determination made by the front determining section 38. In the example of FIG. 19, the front of one building is
It faces a different direction from the front of the surrounding buildings. In such a case, the determination result of the building is corrected based on the front determination result of the surrounding building so that the front faces the same direction as the surrounding building.

In this algorithm, for example, a majority vote is used for correction. Buildings on both sides that are close to the building of interest are required. Here, the distance between the bounding boxes described above is used. Then, the front direction of the building of interest is compared with the front directions of the buildings on both sides. Then, when the frontal directions of the buildings on both sides are the same and only the frontal direction of the building of interest is different, the frontal direction is changed according to the buildings on both sides. This process is performed for each building.

FIG. 22 shows the structure of the texture generator 20. The texture storage unit 50 stores a plurality of types of wall textures, a plurality of types of window textures, and a plurality of types of entrance / exit textures.

The wall texture selection unit 52 selects a wall texture to be attached to the wall surface of the building from a plurality of types of wall textures stored in the texture storage unit 50. Multiple types of wall textures are presented to the user,
One of them is designated by the user. According to this specification,
The wall texture is selected.

The window texture selection unit 54 selects a window texture to be attached to the wall surface of the building from a plurality of types of window textures stored in the texture storage unit 50. Further, the entrance / exit texture selecting unit 56 selects the entrance / exit texture to be attached to the wall surface of the building from the plurality of kinds of entrance / exit textures stored in the texture storage unit 50. The processes of the window texture selection unit 54 and the entrance / exit texture selection unit 56 may be similar to the wall texture, and the texture is selected according to the user's instruction.

The texture generation section 20 carries out a process of pasting the selected texture on the wall surface of the building. The process of pasting the wall texture and the window texture is as already described with reference to FIGS. 15 and 16. The entrance / exit texture pasting process is performed in the same manner as the window texture. However, the entrance / exit texture is attached to a predetermined position in front of the building. The predetermined position is
Typically in the lower center of the front of the building.

The texture sticking process is performed on each of the plurality of buildings included in the map data.

Next, referring back to FIG. 20, the overall operation of the computer graphics processing apparatus 10 will be described. When the user specifies the map data to be processed,
The map data is acquired by the map data acquisition unit 12.
The map data may be created by the user using the computer graphics processing device 10 as described above. The user may specify the range in the map data in which the CG should be created.

Next, when the user specifies the height of each building, height information is acquired by the height information acquisition unit 14. The three-dimensional data generation unit 16 generates three-dimensional data using the building information in the map data and the acquired height information. The front determination unit 18 automatically determines the front of each building using the map data.

Further, the user specifies a wall texture, a window texture and an entrance / exit texture. The wall texture, the window texture and the doorway texture are pasted on the wall of the building according to the user's specification. The doorway texture is attached to the front of the building according to the result of frontal determination.

The CG image with the texture attached is
It is output by the output processing unit 22. For example, the CG image is displayed on the display. At this time, processing such as changing the viewpoint is performed according to the user's request.

The preferred embodiments of the present invention have been described above. According to this embodiment, as described above, the front of the building can be determined by using the distance between the buildings as the determination reference. The front determination result is preferably used for CG image generation. In the above-described embodiment, by attaching the texture of the doorway to the front, an image close to the actual situation can be obtained.

Further, according to this embodiment, a three-dimensional CG image can be obtained by a simple operation of designating the height of a building on the map. In generating a CG image of a cityscape, it is required that many buildings can be easily generated. In this respect,
This embodiment, which can create a CG image by specifying map data and height, is advantageous. However, if the front of the building is not represented,
A feeling of strangeness occurs in the CG image. In particular, in a walk-through image, discomfort may be noticeable. According to this embodiment, the front surface is automatically determined, and a CG image representing the front surface is generated. Therefore, a CG image with less discomfort can be generated. Then, according to the present embodiment, it is possible to provide a technique that enables even an unskilled person to easily create a three-dimensional CG image with less discomfort.

In the present embodiment, the distance between buildings is determined by a relatively simple process by using the bounding box. More specifically, the processing of this embodiment is
Bounding boxes and surround boxes are used to calculate distances using intersections and inclusion points (FIG. 12).

As is apparent from FIG. 12, in the distance calculation of the embodiment, the distance from one bounding box A to another bounding box B is substantially the extension of the side of the other bounding box B. It is the length of the extension line to reach the first bounding box A. Using the length of this extension line as the distance between buildings is advantageous because the calculation process is simple. Then, the processing is simplified by using the intersection points and the inclusion points described in the above embodiment.

Regarding this point, if there is no inclination between the bounding boxes (inclination between sides), calculation of the distance is easy. However, in practice, there is often a tilt between the two bounding boxes, as shown in the examples dealt with in the embodiments above. In this case, in order to consider the front of a building X, the distance from the edge of the face of the building X (the vertex of the bounding box) to another building is
Not appropriate as a judgment factor. The distance from a point on the surface of the building X (a point on the side of the bounding box other than both ends) to another building is appropriate as a criterion. According to the present invention, a distance that is appropriate as a determination factor from the above viewpoint and that is obtained by a simple process is used, whereby the process of frontal determination is appropriately and easily performed.

It should be noted that, within the scope of the present invention, within a range in which the distance between the bounding boxes is appropriately determined, or within a range in which appropriate inclusion points and intersections are determined, a quadrangle other than a rectangle or other polygons, etc. Another form of surround box may be used.

Further, according to the present embodiment, when the intersection cannot be obtained using the bounding box, that is, when there is no other building in the predetermined range around the building,
The largest face of the building is determined to be the front. Therefore, the front can be appropriately determined even in the situation where there is no other building in a predetermined range around the building.

The present invention has been described above based on the embodiments. It should be understood by those skilled in the art that the above-described embodiment is merely an example, and that various modifications can be made to the combinations of the respective constituent elements and the respective processing processes, and that such modifications are within the scope of the present invention. By the way.

For example, in the above-described embodiment, the front determination result is used for the texture generation of the wall surface. The front determination result may be used in another form. For example, the road can be generated based on the front determination result. In an example of the road generation processing, sidewalks having a predetermined width are generated in front of a plurality of buildings. Appropriate modifications are preferably made so that the sidewalk is not jagged. There is a roadway between the sidewalks. Appropriate textures are applied to sidewalks and roadways.

Further, in the above-mentioned embodiment, the map data processing function for making the front face determination is incorporated in the computer graphics processing device. However, the map data processing function may be incorporated in another device. For example, it may be incorporated in a route guidance device using map data. Appropriate guidance that guides the user to the front of the building is possible using the front determination result. Of course, the above-mentioned CG technology may be further incorporated in the route guidance device.

Further, in the above-described embodiment, the determination based on the distance between buildings and the determination based on the size of the surface of the building are combined. However, a configuration in which only one of these is performed is also included in the present invention.

In addition, in the above-described embodiment, FIG.
As described with reference to FIG. 9, the front determination result may be inaccurate. The present embodiment may be configured such that the user appropriately corrects such an error manually. If the front determination result is accurately obtained with a certain high probability, the advantages of the present invention can be sufficiently obtained even if some correction is required. From the same viewpoint, when the fronts of some buildings cannot be determined, such buildings may be presented to the user and the user may be required to specify the fronts. Furthermore, the front surface may be freely changed manually according to the preference of the user.

[0117]

As described above, according to the present invention, it is possible to determine the front of the building in the map data, and it is possible to provide a suitable computer graphic processing technique using the front determination.

[Brief description of drawings]

FIG. 1 is a diagram showing a CG creation system in a study of constructing a three-dimensional cityscape CG image using a digital map.

FIG. 2 is a GeoBase used in the system of FIG.
FIG.

FIG. 3 is a diagram showing a support function of GeoBase.

FIG. 4 is a diagram showing a screen of an application created using GeoBase.

FIG. 5 is a diagram showing a height changing process using a mouse.

FIG. 6 is a diagram showing an example of data exhaled by a GeoBase application.

FIG. 7 is a diagram showing a side polygon generating process.

FIG. 8: G for tiling to generate an upper surface polygon
It is a figure which shows a LU library routine.

FIG. 9 is a diagram showing a bounding box.

FIG. 10 is a flowchart showing a bounding box generation process.

FIG. 11 is a diagram showing a bounding box generation process according to FIG. 10;

FIG. 12 is a diagram showing a process of obtaining an intersection point and an inclusion point using a bounding box in order to obtain a space around a building.

FIG. 13 is a diagram showing a method of creating a rectangle extending around a bounding box.

FIG. 14 is a diagram showing a process of determining an inclusion point.

FIG. 15 is a diagram showing an example of a texture image of a window.

FIG. 16 is a diagram showing a texture pattern changing process.

FIG. 17 is a diagram showing an execution result.

FIG. 18 is a diagram showing an execution result.

FIG. 19 is a diagram showing a front determination result which is an execution result.

FIG. 20 is a diagram showing a configuration of a computer graphics processing device according to an embodiment.

FIG. 21 is a diagram showing a configuration of a front face determination unit in FIG. 20.

22 is a diagram showing a configuration of a texture generation unit in FIG.

[Explanation of symbols]

10 Computer graphics processor 12 Map data acquisition section 14 Height information acquisition unit 16 3D data generator 18 Front judgment section 20 Texture generator 22 Output processing unit 30 bounding box generator 32 Rectangle generator 34 Intersection / inclusion point detector 36 Maximum surface determination unit 38 Front determination section 40 Corrector

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Shiro Fujiwara             7-1 Ogigaoka, Nonoichi-cho, Ishikawa-gun, Ishikawa Prefecture             Kanazawa Institute of Technology F term (reference) 2C032 HC23 HC26                 5B050 BA09 BA17 EA07 EA28 FA02                       FA06

Claims (18)

[Claims] 1. A map data processing method, comprising: a front surface determination process of determining a front surface of a building included in map data based on a distance to surrounding buildings. 2. The surface having the maximum distance to the surrounding building is determined to be the front of the building.
Map data processing method described in. 3. The map data processing method according to claim 1, wherein the front of the building is determined by using the distance between the bounding boxes surrounding the building as the distance between the buildings. 4. The distance from one bounding box to another bounding box is a length of an extension line of an extension of a side of the other bounding box to reach the one bounding box. The map data processing method according to claim 3. 5. A surround box corresponding to each side of the bounding box is formed, an intersection point of the surround box and the bounding box is obtained, an included point which is a vertex of the bounding box included in the surround box is obtained, and an edge of the surround box is obtained. The map data processing method according to claim 3, wherein the distance from the upper intersection to the inclusion point is used for the front determination as the distance between the bounding boxes. 6. The map data processing method according to claim 3, wherein the largest face of the building is determined to be the front when there is no other building around the building. . 7. A map data processing method, characterized in that front view determination processing is performed for determining that the largest face of a building included in the map data is the front face of the building. 8. A step of acquiring map data, a step of judging the front of a building included in the map data based on a distance to a surrounding building, the map data using the judgment result of the front. A computer graphic processing method characterized by generating a computer graphic image. 9. A map data acquisition unit for acquiring map data, and a front determination unit for determining the front of the building included in the map data based on the distance to the surrounding buildings. A computer graphic processing device characterized by using the map data to generate a computer graphic image. 10. The front determination unit determines the front of the building by using a distance between bounding boxes surrounding the building as a distance between buildings.
The computer graphic processing device according to. 11. A map data acquisition unit that acquires map data including a two-dimensional shape and arrangement of a building, a height acquisition unit that acquires information about the height of the building included in the map data, and a map data acquisition unit included in the map data. A front face determination unit that determines the front face of the building based on the distance to the surrounding building obtained from the map data, and obtains the two-dimensional shape up to the height acquired by the height acquisition unit. And a three-dimensional computer graphic image having a three-dimensional shape and a front surface represented according to the determination result of the front determination unit. 12. The computer graphic processing apparatus according to claim 11, wherein the surface determined to be the front of the building is subjected to a process of pasting a texture representing the front of the building. 13. The front surface determination unit determines the front surface of a building by using a distance between bounding boxes surrounding the building as a distance between buildings.
13. The computer graphic processing device according to 1 or 12. 14. A program for causing a computer to execute a process of map data, the computer performing a front face determination process of determining a front face of a building included in the map data based on a distance to a surrounding building. A program characterized by: 15. A program for causing a computer to perform a process of generating a three-dimensional computer graphics image including a building, wherein map data including the building is acquired, and the front of the building included in the map data is surrounded by the surroundings. A program that causes a computer to perform a process of generating a computer graphic image from map data by using the determination result of the front face based on the distance to the building. 16. A computer-readable recording medium in which the program according to claim 14 or 15 is recorded. 17. The front of the building included in the map data,
A map data processing method characterized by performing a frontal determination process of determining based on information on a space up to a surrounding building. 18. The map data processing method according to claim 17, wherein the surface having the largest space up to the surrounding building is determined to be the front surface of the building.